I have a function of an integer and a numpy array. It accepts a single int and a 1D array, or alternatively it accepts an array of ints and a 2D array which are then broadcast, e.g.
def norm_matrix(l, cov):
norm_cov = cov / (l*(l+1))
return norm_cov
My question is how to write the type hints in this case. Does this work?
def norm_matrix(
l: Union[int, np.ndarray],
cov: np.ndarray
):
and is there a better way to do it?
Your solution coul be improved by using numpy.typing.NDarray
From the typing module of numpy
Your way is correct. It is the default way.
If you find it too cumbersome you can do use the bitwise operator | instead as mentioned in the Union documentation:
l: int | np.ndarray
But generally in Python we frown upon mixing input types since it complicates things. But like all rules, this one has an exception: multimethods - Here is an article by Guido explaining how they work.
